Evaluating the Effectiveness of Machine Learning Algorithms in Detecting Distributed Denial of Service Attacks in Mobile Edge Computing

Abstract

Mobile Edge Computing (MEC) is the last mile technology in 5G designed to reduce latency and to execute delay-sensitive applications closer to the end-user. MEC is also known as Multi-Access Edge Computing, which extends the capabilities of cloud computing to the network's edge and brings more capabilities closer to the user. By deploying cloud capabilities to the edge servers, MEC reduces latency and delay. As a result, the end-user experience is enhanced. Unfortunately, the MEC is susceptible to security challenges. Therefore, security is a challenge that requires attention. This paper investigates and compares the effectiveness of machine learning algorithms designed to detect distributed denial of service (DDoS) attacks in MEC. DDoS is one of the security concerns that are negating the advantages of MEC. It degrades the performance of the network. In this work, network and application layer DDoS attacks were considered. The study observed that the Support Vector Machine outperformed other machine learning techniques, and the findings are fundamental to the development of our proposed security framework for MEC.